Primary Production in the Columbia River Estuary I. Spatial and Temporal Variability of Properties!

J. RUBEN LARA-LARA,z BRUCE E. FREY,3,4 AND LAWRENCE F. SMALL3

ABSTRACT: Light , major nutrients, water temperature, turbidity and its or ganic and inorganic fractions, chlorophyll, phaeophytin, DCMU [3-(3,4 dichlorophenyl)-l, I dimethyl ureal-enhanced fluorescence (DCMU ratio), par ticulate organic carbon (POC), particulate organic nitrogen (PON), and primary production were measured from April 1980 through April 1981 in a 65-km stretch ofthe Columbia River estuary. Daily solar input, light attenuation in the water, and chlorophyll concentration accounted for 75% of the variability of daily primary production in the main estuarine axis and 85% in the shallows. The rapid appearance of a turbidity load created by the Mt. Saint Helens volcanic eruption in May 1980and the subsequent clearing of the water as the load moved out ofthe estuary became a natural experiment to show that light availability was indeed the limiting factor to phytoplankton production in the estuary. Spatial variability in chlorophyll concentration was caused mainly by large summer reductions at the location where freshwater cells were lysed on contact with low salinity intrusions. Mean values for properties in the main axis generally were not significantly different from those in the shallows, suggesting that the main axis and shallows experience similar , rapid flushing times. Total primary production for the estuary was almost 30,000 metric tons C yr", but areal production was only 100 g C m- 2 yr" , which puts the Columbia system at the low end ofNorth American estuaries. The low areal production was likely a result of light limita tion, chlorophyll reduction at the low-salinity boundary, and a short residence time of water and viable cells in the estuary.

OF ALL THE aquatic ecosystems, estuarine sys adapt. Factors vary in response to tidal, diel, tems are perhaps the ones exhibiting the widest and seasonal cycles, as well as to sporadic and most frequent physicochemical fluctua changes caused by storms and mankind's in tions and the ones to which phytoplankton tervention. For these reasons, understanding (and other organisms) find it most difficult to of the mechanisms controlling the estuarine abundance of phytoplankton, its production, and the species composition ofpopulations is most difficult. Yet estuaries are of primary I This study was supported by a grant from the Pacific Northwest River Basins Commission and the U.S. Water importance both from an ecological and an Resources Council, administered through the Columbia economic view.The lower Columbia River is a River Estuary Data Development Program (CREDDP) coastal plain estuary, or drowned river valley, and Columbia River Estuary Study Taskforce (CREST) . separating the states of Washington and J. R. Lara-Lara had a scholarship from the National Oregon in northwestern USA. The estuary Council of Science and Technology in Mexico. Manu script accepted 12 May 1989. was formed as the sea level rose to its present 2Centro de Investigaci6n Cientifica y Educacion Su position after the last glaciation. However, the perior de Ensenada , Ave. Espinoza no. 843, Ensenada, Columbia River estuary is much more river B. c., Mexico 22830. dominated than most other coastal plain estu 3College of Oceanography, Oregon State University, Corvallis, Oregon 97331. aries in the world. The Columbia River is the 4 Present address: Oregon Health Sciences University, second longest river in North America, with Portland, Oregon 97201. the second largest volume of discharge in 17 18 PACIFIC SCIENCE, Volume 44, January 1990

FIGURE I. Map of the study area. Estuarine regions (numbers) are delineated by hydrographic and sedimentary properties. Zones (letters) are (a) marine zone. (b) estuarine mixing zone, and (c) freshwater, or tidal fluvial. zone. the United States (annual discharge is about mainly on river hydrology, circulatory pro 2 x 1011 m - 3 , or 58% that of the Mississippi cesses, and sedimentary geology, Simenstad et River). The Columbia and its tributaries drain al. (in press) sectioned the estuary into three 2 a region of ca. 660,500 m • River flow is now principal zones and 10 different regions for greatly regulated by dams, but still varies from purposes of analysis along the estuarine con about 3,000 to 12,000 m' sec-I. tinuum and in the adjacent bays. For our Williams and Scott (1962) and L. G . Williams phytoplankton studies, confined to the water (1964, 1972) reported the phytoplankton flora column, we could not distinguish between re of the estuary. Haertel and Osterberg (1967) gions 3 and 5 in the estuarine mixing zone, and and Haertel et al. (1969) described some as therefore treated that area as combined region pects of hydrography and plankton ecology (3 + 5). for this system, and Park et al. (1969, 1972) Nine cruises were conducted approximately discussed the estuarine nutrient budget. How every other month from April 1980 to July ever, before our research there had been no 1981. Stations in both shallows and main reports on phytoplankton production for this channels were sampled (Figure 2). With ex ecosystem, and no attempts to understand the ception of the June and July 1981 cruises, mechanisms that control the phytoplankton when only three stations were sampled, the biomass in the estuary. Here we report the number of primary production experiments spatial and temporal distribution of phyto per crui se varied from 5 to 12. To study the plankton biomass and primary production as spatial and temporal variability of tempera well as the distributions of the physicochemi ture, selected inorganic nutrients, particulate cal driving variables in the estuary throughout chlorophyll a and phaeophytin a, particulate an annual cycle. organic carbon and nitrogen, total suspended particle load, and the organic and inorganic fractions ofthe total suspended load, from 25 to 47 stations were sampled on each cruise. MATERIALS AND METHODS All stations were sampled at the surface, and The study area extended from the mouth of at the deeper stations depths of 2.5, 5.0, and the estuary 56 km inland (Figure 1). Based 10.0 m were sampled with Van Dorn bottles. Primary Production in the Columbia River Estuary I -LARA-LARA ET AL. 19

FIGURE 2. Sampling stations for the investigation of water column primary production. 0 = biomass measurements only; • = primary production and biomass measurements. The cross-hatched areas are those containing the shallow water stations .

Primary production was assessed by the puted from alkalinity measurements made by radiocarbon uptake method (Strickland and potentiometric titration with 0.1 N H 2S04 , as Parsons 1972). Because the shallow euphotic described by Wetzel and Likens (1979). zone was always well mixed, samples were all Because phytoplankton was distributed collected from the surface. Once collected, the homogeneously through the photic zone, and samples were labeled with carbon-14 and in because the light intensity production rate cubated in 80-ml clear polycarbonate bottles, relationship was unchanging over the ex with two replicates for each incubation. Incu perimental time (Lara-Lara 1982), conversion bations usually were done for 4 hr around of primary production during the incuba local noon under natural sunlight in clear deck tion period to daily integral production 1 tanks. Surface water was circulated through (mgC m-2 d- ) was reasonably straightforward the tanks to maintain temperature. Except for (Vollenweider 1971). Basically, a "surface rate surface samples, light was attenuated with curve" was determined from an equation neutral screens to 50, 30, 15,6, 1, and 0% of describing the photosynthesis: depth curve incident light. At the end of all incubations, (Vollenweider 1971) and a curve ofdaily pho the radiolabeled samples were immediately fil tosynthetically active radiation (PAR). The tered through 0.8-/lm-pore-size membrane fil ratio of the area under this surface rate curve ters, and the filters were preserved in Aquasol to that fraction of the curve during which in individual liquid scintillation vials. No sig incubations were made was used to convert to 1 nificant differences were found when we com mg C m- 2 d- . Specification of the photosyn pared fumed (HCl) and unfumed carbon-14 thesis: depth curve and daily PAR curve re samples. This was expected, because the estu quired measurement of PAR at the estuary arine phytoplankton was mainly diatoms and surface (10 ) and determination of the diffuse microflagellates (Amspoker and McIntire light extinction coefficient (k). 1986). Radioactivity was analyzed in a Beck Daily PAR (295-695 nm) was measured by man Model 7500 liquid scintillation counter. an Epply precision recording pyranometer Carbonate alkalinity, for conversion of radio mounted at a land-based location about activity to carbon-based production, was com- 30 km up-estuary from the mouth. Light pen- 20 PACIFIC SCIENCE, Volume 44, January 1990 etration into the water column, for determina data (Jay and Smith [in press]; Sherwood and tion ofk and the photic depth, was measured Creager [in press]) were selected to represent with a Licor submersible spherical quantum distributions ofproperties along the main axis meter. of the estuary, from near the mouth through Chlorophyll a and phaeophytin a analyses, the most riverine station at the head of the to determine the vertical distribution of auto study area (Figure 2). Other'stations were used trophic phytoplankton through the euphotic to show distributions of properties in the major zone as well as spatial and seasonal distribu shallow bays (Youngs, Grays, and Cathlamet tion, were done by fluorometric measurement bays) and tributary rivers (Youngs, Lewis and on acetone extracts of particles of 0.8-jim fil Clark, and Deep rivers). ters (Strickland and Parsons 1972). In addi tion , in vivo fluorescence measurements were Properties along the Main Axis ofthe Estuary made both with and without the electron transport block DCMU [3-(3,4-dichloro Surface temperature (Figure 3, top) fol phenyl)-l, 1dimethyl urea], at all stations and lowed the solar irradiation cycle, as expected; depths sampled. The ratio between fluore thus, temperature increased from 9° to 15°C scence after DCMU treatment and fluores during spring (April-May), peaked in summer cence before treatment was called the DCMU (at 22'soC), decreased to 10°C by late fall, and ratio, and this ratio was used as a rough mea reached its minimum in winter (5-6°C). sure of the photosynthetic capacity ofthe phy Haertel et al. (1969) and Park et al. (1972) toplankton; that is, the greater the ratio, the reported a similar cycle for the estuary. The greater the photosynthetic capacity (Samuel extinction coefficient oflight (k) in the estuary son and Oquist 1977, Vincent 1981). Samples was a function of the turbidity of the water were also taken for the determination ofpar (Figure 3, bottom). On 18 May 1980 the vol ticulate organic carbon (PaC) and nitrogen canic eruption ofMt. Saint Helens caused the (PaN) by Perkin-Elmer 240C elemental ana discharge of an exceptionally high load of lyzer. Water temperatures were read from a sediment and detritus into the Columbia River thermometer submerged in a bucket filled above our estuarine study area. Extinction co with water from each collection depth. Salini efficients exceeded 6.0 m-1 on our sampling ties were measured to the nearest part per date a few days after the eruption and were thousand with a Goldberg TIC refractometer. still above 4.0 in mid-June (Figure 3, bottom). Detailed temperature and salinity data at se By July, however, k values were back into the lected times and locations were measured by range ofpre-eruption values. Discounting the Jay and Smith (in press), but seasonal patterns period affected by the volcano, k varied from were only possible to elucidate with our own ca. 0.8 to 3.5 throughout the year. Greatest data. Inorganic nutrients (phosphate, silicate, light extinction occurred in the fluvial and and nitrate + nitrite) were analyzed using a mixing zones of the estuary in late summer Technicon Autoanalyzer, according to the and fall [regions 8, 7, and (3 + 5) in Figure 1], techniques of Atlas et al. (1971). Total sus while greatest light penetration into the water pended particles (TSP), and the organic (aSP) column occurred generally in spring and near and inorganic (ISP) fractions ofthe total, were the estuary mouth. Extinction coefficients determined by gravimetric analysis before and between 0.8 and 3.5 equate to photic depths after peroxide oxidation (Peterson 1977). (depths at which surface radiation is reduced to 1%) between 1.3and 5.8 m. A large portion of the estuary had a photic depth of < 2.5 m over most of the year. RESULTS Dissolved phosphate (Figure 4, top) was Stations with water depths ~ 10 m in generally higher in winter and early spring regions 1, (3 + 5), 7, and 8, or with depths (0.8-1.0 jiM) than in summer and early fall < 10 m but part ofthe main estuarine contin (0.3-0.8 jiM). Nitrate plus nitrite (Figure 4, uum from hydrographic and sedimentological bottom) also registered maximum values in Primary Production in the Columbia River Estuary I -LARA-LARA ET AL. 21

~ 151 I o I 70 II) 20 6 i 20 1 1ii 253 o 60 II) ~ I .... 8 301 o 50 (/) z 10 (/) o 8 a: 40 ~ ~ g' 405 o lLJ I- .;;C ~ (/) ·E 4 5 1 30 3 0'10 14 ~ 20 ~ 501 . ~ 552 o :) 10 E 551 J

L..-_L..---lL..---l'-----'_---'_--L_----L_-'-_-J...._-J...._--L._--L.-.I0 AM JJ A SON o J FMA 19 80 1981

~ 151 II) 70 2 201 i ,~ o \ 1ii 253 / , /0 1 , "" 60 ~ ~21 ," 3, 2 30 1 0 4 I'0 •" \ 50 (/) 2 \ z 26 2 <2 o " \ >3 \I \ , ,, ~ g' 405 " I I- .;;C , " (/) ·E 451 :\~ ".. _0.. , • 20 ~ 501 4t 1 • . ~ 552 I, 8 e 0 , 10 E 551 o '

AM J J A SO N o J FMAM 19 80 198 1 FIGURE 3. Spatial-temporal distributions of (top) surface temperature CC), and (bottom) light extinction coefficient, k (m"). Dash ed lines represent a change in contour interv al, and thin lines represent a lower level of confidence in cont ouring due to a reduced number of stations. The vertical lines isolating Ma y and part of June 1980 (bottom) delineate the time over which the volcanic eruption had significant effect. winter and early spring (> 20 jlM), while min early spring to lower values (around 100 jlM) imum values during summer were < 1.0 jlM. in summer. Concentrations of silicic acid in No other nitrogenous nutrients besides nitrate fall and winter dropped abruptly near the es and nitrite (ammonia, for example) were mea tuary mouth because of intrusion of coastal sured in the estuary, so we cannot say for sure oceanic waters, which contained much less that nitrogen was absent or nearly absent dur silicic acid than river waters. ing summer. Silicic acid (Figure 5, top) varied The distribution of total suspended parti from high values (160-200 jlM) in winter and cles (TSP) showed an abnormally high con- 22 PACIFI C SCIENCE, Volume 44, January 1990

~ 151 o 70 II) \ ~ 201 0.6-....:. ~ 253 \ o 60 CD ~ 0.8 -301 50 en en z a: 0 40 ~ t- c:{ ~ 405 • UJ t- .;:c :::E en ·E 451 • O·ti 30 g ~

~ 501 • .;: 552 c ,96d/: E 551 Oil 5 0 AM JJ A S 0 N 0 J FM A 1980 1981

~ 151 , "\ o 70 II) 3·, , 25 , ~ 201 0, , oJ ~ ~ , " ,. " , 253 ., , ,'.-, , OJ , , 60 CD ~ I 1<1' I I I I \ 20 301 I , ' r, , , - I 50 0' ," , I I I I I en I I en Z I \ J , I I a: 0 , I / , 10 I , 'I" I I 40 ~ I I t- I , UJ c:{ ~ 405 , I I I .;:c ., . I· :::E t- I ...... \ I en ·E 451 . , l • 30 g , ·...... '--' · ~ 30 5 5 I ,-, ... ~ 501 .; I " 'c 552 , ,.'

0 A M JJ A S 0 N 0 J F M A 1980 1981

FIGURE 4. Spatial-tempo ral distributions of (top) surface-dissolved phosphate (PM), and (bottom) surface-dissolved nitrate (PM). Other comments as in Figure 3. centration peak during May 1980 as a result of and minimum ranges for ISP were 45-70 and 3 the Mt. Saint Helens eruption (Figure 5, 7- 15 g m- , respectively, while asp maxima bottom). Other than that, maximum loads and minima were 4- 8 and 1- 3 g m'. Typical were in late summer and fall (50-70 g m- 3), concentrations for particulate organic carbon with minimum concentrations in midwinter and nitrogen were from 1.0 to 1.3 g m-3 for 3 3 and early spring (8-20 g m- ) . The inorganic pa c and from 0.1 to 0.2 g m- for paN (not (lSP) and organic (a SP) fract ions of the TSP illustrated). Slightly higher values were regis (not illustrated) mirrored very closely the pat tered during midwinter and early spring. tern of total suspended particles. Maximum The phytoplanktonic chlorophyll a concen- Primary Production in the Columbia River Estuary I -LARA-LARA ET AL. 23

.. 15 1 ., 70 CD , C , ~ · >",.~, \ I. 201 • ~ 253 ;' \.... 1; 60 CD · .- .. .. 190 120 " , 180 - 301 .' . , 50 (/) 18 I/O (/) z , a:: 0 , 40 ~ t- , « ~ 405 • 40 ., W t- ';; 1.70 I ~ (/) "e 451 .\ • 30 ::: \ , , ~ •I 20 ~ 501 • . 160 , .;: 552 • I .' 0 10 E 551 ~

0 A M J J A S 0 N 0 J F M A 1980 19 8 1

.. 15 1 ., . 70 CD '"460 , I 30 ~ 2 01 . ~ · I 253 • I ~ 60 CD 10 .. I 00 20 3 01 - •II (/) I Z 0 I I t- ~ 405 I « • I t- ';; I (/) 'e 4 51 • I I I I ~ 5 01 •I .;: 552 • I 0 I E 551 .,

0 A M JJ A S 0 ND J F M A 1980 19 81

F IGURE 5. Spatial-temporal distributions of (top) surface silicic acid (JIM), and (bottom) surface total suspended particles, TSP (g m"). Other comments as in Figure 3.

tration and the DCMU ratio varied both fluvial stretch of the estuary to < 2 mg m - 3 in spatially and temporally (Figure 6, top and the entrance region. During spring and sum bottom). In a plot ofmean chlorophyll a con mer the mean concentrations usually showed centrations by sampling months and by re a striking decrease at the interface between the gions for the main body of the estuary, the tidal fluvial zone and the mixing zone. In Ju ly spatial and temporal variability became more 1980, for example, concentrations fell from an evident (Figure 7). During fall and winter, average of 14.1 mg m- 3 in the main-axis part for example, mean concentrations decreased of region 7 to 6.3 mg m- 3 in the adjacent rather uniformly from > 5 mg m- 3 in the region (3 + 5) (Figure 7). Concentrations in 24 PACIFIC SCIENCE, Volume 44, January 1990

... 151 CD •I j 201 .6 ~ 253 ...CD - 301 50 m z 0 I-

20 /',/, I. \ '.I\, 10

0 A M JJ A S 0 N 0 J F M A 1980 1981

151 • 2.5 70 2. :2 (:",2.7"': :§ 201 . \y/" " ;; 253 60 2.5 "" ... ) "', 301 '" 50 m m z Q a: 0 - 40 ~ I- 2 10 o~551... · .. .. c -CIl 0 A M JJ A S 0 N 0 J F M A 1980 1981

FIGURE 6. Spatial-temporal distributions of(top) surface chlorophyll a (mg m- 3), and (bottom) the ratio ofDCMU induced fluorescence to in vivo fluorescence. Other comments as in Figure 3. region 8 always were similar to those in region are also transition months. Even though the 7 in spring and summer, while concentrations May 1980 chlorophyll a concentrations re in region 1 were reasonably similar to those in flected the eruption of Mt. Saint Helens to region (3 + 5). Distributions in April 1980 some unknown extent, the distinct decline in were more winterlike, while distributions in chlorophyll a between regions 7 and (3 + 5) April 1981 were closer to those found in sum was still evident at this time. Phaeophytin a mer, probably attesting to the spring months (not illustrated) showed maximum values in 3 as being months of transition between winter late summer (2.0-3.5 mg m- ) and minimum and summer conditions. Likely the early fall concentrations during midwinter and early 3 months (September in Figure 7, for example) spring (0.0-0.7 mg m- ) . Primary Production in the Columbia River Estuary I-LARA-LARA ET AL. 25

If) -IE 20 0' oS 7 8 >- 10 3 s: 5 0- I 0 ~ 0 s: 0 U Apr. May July Sept. Noy. Feb. Ap r. 1980 1981

3 FIGURE 7. Mean chlorophyll a concentrations (mg m- ) by sampling months and by regions (see Figure I) along the main axis of the Columbia River estuary. May 1980 concentrations reflect the eruption ofMt. Saint Helens to some unknown extent. Regions 3 and 5 are combined because they were not differentiated during samp ling for chlorophyll.

Like chlorophyll a, the DCMU ratio was nitrate plus nitrite (> 3.0 /lM) and phosphate significantly variable both temporally and (> 1.0 /lM) relative to riverine waters (which spatially (Figure 6, bottom). With the excep normally ranged below 1.0 /lM of nitrate plus tion ofMay 1980, when the volcanic eruption nitrite and 0.5 /lM phosphate) (Figure 8). tended to decrease the DCMU ratio even in From late fall through early spring, riverine the fluvial regions of the estuary (Figure 6, waters were highly enriched with respect to bottom), the spring and summer months nitrate plus nitrite and to a modest amount yielded enhanced ratios. In summer the highest with respect to phosphate. In contrast, silicic ratios were in the fluvial stretch, whilein spring acid concentrations were always higher in the the highest ratios often extended from the river waters than in marine waters (Figure riverine areas through part ofthe mixing zone. 8). Marine waters showed maximum silicic The lowest ratios tended to occur in that part acid concentrations during summer upwelling of the mixing zone adjacent to the entrance (close to 100 /lM) , when river waters showed region, regardless of season. The entrance re their minimum (slightly greater than 100/lM). gion itself often supported DCMU ratios> 2.0 During fall and winter, river waters ranged throughout the year. from about 120 to > 180 /lM, while enter Except for the entrance region and the west ing marine waters were generally <40 /lM. ern part of the mixing zone, vertical distribu Chlorophyll a concentrations were nearly al tions of particles and most physicochemical ways higher in river-derived waters than in properties along the main axis of the estuary marine waters in these regions of substantial were reasonably homogeneous. During peri mixing (Figure 8). ods of low river flow in the summer, tempera For purposes of estimating phytoplankton ture and salinity often showed distinct vertical production over the entire estuary rather than inhomogeneity well into the tidal-fluvial zone just at the stations where carbon-14 measure .(Jay and Smith [in press]). Vertical distribu ments were made , empirical models based on tions of six properties to 10m depth during the relationships between measured produc flood tide in four different months are shown tion rates and 10 ecological variables were in Figure 8 for three stations in the entrance developed separately for the shallow and deep and south channel regions. The vertically stretches of the estuary. Step-wise multiple stratified water column in July with respect to regression analysis (Rowe and Brenne 1981) temperature and salinity is obvious, but in the was used to develop the relatio nships, with other months the stratification is more nearly daily solar input, the light extinction coeffi horizontal even at these near-ocean stations. cient, water temperature, chlorophyll a, During summer, marine coastal upwelled phaeophytin a, nitrate plus nitrite, phosphate, waters entered the estuary slightly enriched in silicate, total suspended seston load, and the 26 PACIFIC SCIENCE, Volume 44, January 1990

Stations 552 552 552 1591 5 91 1591 5~1 1 5q l ~ • ./1.5 ~ .~/I" .O . I I 1 1 I I '1. II . ·17.3 '.10.. 15~ ·12.8 -E " '\ 20 5. 10. 7.9 I II III o.7 b~~1 ·1.0 EJ0.9 10 III III I I I 0..-:----:-----::--1 ·165 5 ~~/IO ~0~40 0//2 Si(OH)4 700 ~"; 30 10L....t._---;----;---' (JLM)

1 I 1 710 5 III I I O :'~ '7. 6 ~ 5 r-5 • • ~4 10 Vf' July Nov. Feb. April

FIGURE 8. Vertical distributions of properties in the near ocean and lower estuarine mixing zones. All data were taken on flood tide. organic portion of the total suspended seston Table I. The ranges are broad, however , and being the variables examined. Five of the 10 encompass the ranges characteristically found variables cumulatively accounted for 90% of over the year in the estuary. More elegant the variability in primary production in the theoretical models might extend the limits of main axis ofthe estuary (Table I). Solar radi prediction of primary production somewhat ation input and the light extinction coefficient and will be examined in future work . Using alone accounted for 75% ofthe variability. It both measured and calculated values, daily must be emphasized that the model in Table I phytoplankton production showed strong is specific to the Columbia River estuary and seasonal variation (Figure 9), as expected from can predict primary production only within the annual patterns of solar radiation and the ranges ofthe fivekey variables indicated in chlorophyll (Figure 7). Unlike chlorophyll Primary Production in the Columbia River Estuary I - LARA-LARA ET AL. 27

TABLE 1 PRIMARY PRODUCTION REGRESSION MODEL FOR STATIONS IN THE MAINAxis OFTHE COLUMBIA RIVER ESTUARY (n = 29)

ABBREV. VARIABLE CUMULATIVE R2 MODEL

S Daily solar radiation (g cal cm ? day ") 0.58 Log dai ly production = 1.548 + O.OOIS - 0.103k + 0.56Chla + 0.28T - O.OOITSP k Light extinction coefficient (m - ') 0.75 3 Chla Ch lorophyll a (mg m- ) 0.84 T Temperature eC) 0.87 TSP Total seston (g m-3 ) 0.90

NOTE: Lower and upper ranges of values for the above variables: S = 250-4,800 k = 1.06-9.29 Chla = 0.7-17.9 T = 5.1-21.3 TSP = 5.8-84.7

'->- C "0 a a N 1000 'E o 7 CI 3 E BOO 7 5

~ 0 600 +-o :;, 3 "0 400 0 5 a.~ 7a >- 200 ~ c E 0 a.~ Ap r. May July Sep t. Nov. Feb. Apr. 19 80 1981

FIGURE 9. Mean daily phytoplankton production (mg C m ? day") by sampling months and by regions (see Figure I) along the main axis ofthe Co lumbia River estuary. Actual rates in May 1980 (hatched bars) are reduced because of the Mt. Saint Helens eruption. Rates without the volcano effect (open plus hatched bars) have been computed and are also shown for May 1980. Numbers indicate the estuarine regions from Figure I.

concentrations, however, primary production in the absence of the volcanic eruption. Re was greatly affected by the volcanic eruption. calculated estimates of May production rates By reducing light penetration to a great extent were close to those for July in the riverine in May 1980, volcanic debris in the estuary stretch of the estuary and more nearly repre reduced primary production in all parts ofthe sentative of rates expected in spring (Figure main estuarine axis by 75% (Frey et al. 1983). 9). Large differences among the four regions Correction of the measured May values was along the estuary axis were observed only in done by changing the light extinction coeffi May and July. The rather precipitous decreases cients in the model equation to values expected in chlorophyll between regions 7 and (3 + 5) 28 PACIFIC SCIENCE, Volume 44, January 1990 during May and July (Figure 7) were in large Properties in the Shallows and Bays ~art responsible for the decreases in produc Temporal variability of properties in shal tion between the same two regions; however, low areas showed the same seasonal trends as close similarities in chlorophyll concentration the comparable properties along the main axis between regions 8 and 7, and between regions (Figure 11). (~ 5) and 1were not matched in the produc + Daily integral primary production in the tion data. Furthermore, no trends in the pro shallows was adequately modeled using only duction data by region were noted in the other daily radiation input, light extinction in the months (Figure 9). Such dissimilarities pointed water column, and chlorophyll a concentra up the fact that light mainly controlled pro tion as variables (Table 2). The pattern of duction, with only moderate effect from chlo production developed from model-generated rophyll concentration, on average (Table 1). estimates plus measured rates, by region, Selected vertical distributions of measured showed strong seasonality as expected (Figure primary production for representative stations 12). What was not necessarily expected , how along the axis of the estuary are shown in ever, was the lack of significant difference Figure 10. The typical seasonal pattern was (P> 0.05) between mean daily production at evident , with maximum hourly rates and the shallow stations and the main-axis stations deepest penetration of production in summer (Table 3). The possibly longer residence times to yield highest integral production at that ofcells in the shallower, more isolated reaches time of year. Lowest integral production oc of the estuary, given sufficient nutrients and curred in winter, the result of reduced solar light throughout the shallow water columns input and a restricted euphotic zone, and re suggested that the shallows might have bee~ duced chlorophyll concentrations. more productive on an areal basis. Neither the mean chlorophyll a nor the light extinction coefficient by seasons were significantly dif ferent when sta tions in the main axis were compared with stations in the shallows (Table 3). The mean assimilation ratio was only signi fi~antly hi~her in the shallow regions during wmter--sprmg 1980 (Table 3). The lack ofsig nificant differences between the means sug gests that mean residence times ofcells in the photic zones of both areas are similar.

Estuary- Wide Primary Production The production data from Figures 9 and 12 (excluding the tributary rivers) can be inte grated through one annual cycle after remov ing the effects ofthe volcanic eruption to yield annual areal production for each estuarine region (Table 4). The decrease in areal pro duction from fluvial to entrance region down the main axis [regions 8, 7, (3 + 5), 1]is easily observed. The more isolated bays have the lowest areal production, although the lack of SI. 501 SI.451 SI. 201 production data in the very shallow Baker and Trestle bays (region 2) makes the estimate .F IGURE 10. Typical primary productivity profiles near midday for three selected stations during four different there little more than speculation. seasons. When the surface areas of each region in

... . W '" • Primary Pr oduction in the Columbia River Estu ary I-LARA-LARA ET Ai. 29

U 20 ~ o - - .:!-- 1.0 Q) o ..c- 0.5 Q. eI) o ..c 0.

-'-~ 200 -~ .:!-- ~ 40 " . "0 150 Q) o 50 - .I' \' \/' ~ , \ -, /\ , ... "'/."" u - ...... / ....\ ' '\ o 25 0:':'. A J A o AFD 0:,:",,:••• oL.-...... '--'---'--'--'--'---'-"--J...... --L--'--' A J A o o F A s: Ie c> 20 t- 1\ Youngs Bay e : -. I\ Youngs and Lewis Clark Rivers .I: "". \ a 01 15 : 1-.-.-:.... : t. .. Deep River >0 10 :/ s: " Q. , Grays Bay e 5 o Cathlamet Bay ..c OL.-L...... JL...... JL.-L...... JL...... JL...... JL...... J'--L.-L...... JL...... J'--...... U A J A o D F A

FIGURE II. Temporal variability ofproperties in shallow areas (see Figu re 2). Values are for surface samples, which reflect values through the shallow water columns in the different regions, 30 PACIFI C SCIENCE, Volume 44, January 1990

TABLE 2 PRIMARY PRODUCTION REGRESSION MODEL FOR SHALLOW STATIONS AND BAYS IN TIlECOLUMBIA RIVER EsTuARY(n = 28)

ABBREV. VARIABLE CUMULATIVE R2 MODEL

S Daily solar radiation (g cal em"? day") 0.73 Log da ily production = 1.605 + 0.003S + 0.033Chla - 0.127k k Light extinction coefficient (m- I) 0.80 Chla Chlorophyll a (mg m") 0.85

NOTE: Lower and upper ranges of values for the above variables: S = 250-4,800 k = 1.26-8.10 Chla = 0.8-24.0

I>. c 9 '0 10 0 0 N IE 7 u C' 800 9 ..s 7 c: 600 0 6 4 +- o 4 10 ~ 4 9 '0 400 0 9 ~ 6 a. 7 6 7 200 >. 6 10 ~ c 4 46 7 9 10 467910 E 0 July Apr. ~ Apr. May Sept. Nov. Feb. a. 1980 1981

FIGURE 12. Mean dail y phytoplankton production (mg C m-2 day") by sampling months and regions in shallow water bays and tributaries ofthe Columbia River estuary. Actual rates in May 1980 (hatched bars) are reduced in the bays because ofthe Mt. Saint Helens eruption, but the tributary rivers were not affected. Rates in the bays without the volcano effect (open plus hatched bars) have been computed and are also shown for May 1980. Numbers indicate the estuarine regions from Figure 1.

Figure 1 are computed and multiplied by the DISCUSSION annual areal pro duction rates for each region, estimates oftotal primary production are gen In general, light and the availability of erated for each region (Table 4). We divided nutrients are the enviro nmental factors that region 7 into a shallows area and a main-axis regulate phytoplankton prod uction in most area (Figure 2) and computed surface areas marine and estuarine systems (Ryther and and productio n rates separately for these two Dunstan 1971, MacIsaac and Dugdale 1972, areas. Region (3 + 5), by virtue of its large R. B. Williams 1973, Ball and Arthur 1979, area, had the greatest total ann ual phyto Sharp et al. 1982). Inorganic nutrient supply plankton production, approximately one-third seems to exert little control on primary pro of the total estuarine pro duction ofabout 3 x duction in the Columbia River estuary, 104 metric tons of carbon per year. however. Altho ugh all measured nutrients de-

~. Primary Production in the Columbia River Estuary I-LARA-LARA ET AL. 31

TABLE 3 COMPARISONS OFPROPERTIES FOR MAIN Axis AND SHALLOW STATIONS IN THE COLUMBIA RIVER ESTUARY

MAIN AXIS SHALLOWS S VARIABLE SEASON MEAN ± SE MEAN ± SE (P < 0.05)

Primary productivity Winter-spring 1980 216.8 (35.80) 257 .7 (94 .20) ns (mg C m - 2 day "") Summer-fall 1980 396.4 (134.40) 472 .3 (112 .70) ns Winter-spring 1981 200.3 (81.70) 212 .0 (83.00) ns Chlorophyll a Winter-spring 1980 10.3 (1.84) 8.0 (3.10) ns (mg m - 3 ) Summer-fall 1980 7.9 (1.40) 9.3 (1.70) ns Winter-spring 1981 6.6 (1.00) 9.3 (3.70) ns Assimilation ratio Winter-spring 1980 1.25 (0.17) 4.22 (1.18) sig. i i (mg C mg Ch1a- h- ) Summer-fall 1980 3.40 (0.55) 2.78 (0.27) ns Winter-apring 1981 1.71 (0.28) 1.40 (0.28) ns Light ext inction Winter-spring 1980 4.15 (1.3 1) 3.31 (1.01) ns coefficient k Summer-fall 1980 2.08 (0.26) 2.46 (0.23) ns (m"") Winter-spring 1981 1.76 (0.17) 2.35 (0.33) ns

NOTE: Means represent data from all cruises ± I SE.

TABLE 4 ANNUAL RATES OFNET AREAL AND TOTAL PHYTOPLANKTON PRODUCTION BY REGION

AREAL PRODUCTION TOTAL PRODUCTION i i REGION (g C m? yr - ) (metric tons C yr - )

1 Entrance region 80.3 2,493 .3 2 Baker Bay and T restle Bay (72. 1) (1,192 .6) 3+5 Estuarine channels and mid-estuary shoals 84.9 10,638 .8 4 Youngs Bay 63.9 816.2 6 Grays Bay 66.3 2,323 .5 7 Cathlamet Bay (shallows) 102.8 3,102.5 7 Cathlamet Bay (main axis) 146.2 4,412.3 8 Fluvial region 152.7 4,891.7 Mean (including region 2) 96.2 Mean (excluding region 2) 99.6 Total (including region 2) 29,875.9 Total (excluding region 2) 28,683.3

NOTE: The areal estimate for region 2, which was not sampled, was taken as a mean of the annual rates for regions I and 4. creased in concentration in summer, at least during summer upwelling, for example. Dur partly because of greater phytoplankton de ing our summer sampling, we also observed mand, concentrations never became too low an enrichment of nitrate (and phosphate) in to measure, nor were they strongly correlated waters ofregion 1 deeper than 5 m (Figure 8); with primary production (Tables 1 and 2). however, at no time did we measure nitrate Similar nutrient cycles have been reported concentrations in the summer as high as those earlier for the Columbia River estuary by observed by Haertel et al. (1969). The inten Haertel et al. (1969) and Park et al. (1972). sity and extent of this summer enrichment Part of the control on nutrien t levels in the must depend main ly on the timing and inten estuary is through offshore upwelling. Haertel sity of the coastal upwelling events. Halpern et al. (1969) observed nitrate concentrations in (1976), Walsh et al. (1977), and Small and the entering marine waters of up to 23 ,uM Menzies (1981) indicated that the duration of 32 PACIFIC SCIENCE, Volume 44, January 1990 upwelling events in the eastern Pacific, includ which coincided with the re-establishment of ing off the northwestern coast of the United light extinction values typical ofsummer (Frey States in summer, was anywhere from a few et al. 1983). The Mt. Saint Helens eruption days to several weeks, interspersed with non thus performed a large-scale experiment in upwelling conditions of a few days' duration. which light penetration into the estuary was Some events are intense , with nitrate levels of rapidly and drastically reduced without con 30 ~M surfacing near shore, and some are comitant reduction in nutrients or chlorophyll relatively weak with concentrations of 10 JlM a. The reduction and subsequent recovery of inshore (Small and Menzies 1981). Such spo primary production was directly caused by the radic inputs from the ocean , plus variable reduction and subsequent re-establishment of inputs from the riverine end of the estuary the photic depth in the estuary. Sharp et al. through the year, undoubtedly erode any rea (1982) postulated severe light limitation in the sonable correlation between point estimates of upper Delaware estuary, USA, as a result of nutrient concentration and point estimates of high suspended sediment concentrations, and primary production. In addition, not all ni Pennock (1985) indicated that chlorophyll trogenous nutrients were measured (ammonia, distributions in the Delaware estuary were for example), further suggesting that nutrient light-limited. supply did not exert major control on primary Salinity exerted some control on primary production. production, and on chlorophyll biomass, in a Light is the main variable controlling phy specific way. Chlorophyll a and production toplankton production in the Columbia River along the main axis almost always decreased estuary. Cumulative multiple regression anal from the fluvial to the entrance regions (Fig ysis showed thatdaily solar radiation input and ures 7 and 9), but dramatic reductions only light extinction in the water column accounted occurred in spring and summer between re for 75% ofthe variability in production along gions 7 and (3 + 5). Haertel et al. (1969) also the main estuarine axis (Table 1) and 80% in reported that cell numbers as well as chloro the shallows (Table 2). In addition, the vol phyll a decreased downstream as the salinity canic ash and mud from Mt. Saint Helens increased , with the most striking changes oc reduced primary production immediately by curring in the "mixing zone." We plotted the reducing the photic zone in the estuary (Fig occurrence of numerically dominant fresh ures 9 and 12; Frey et al. 1983). Nutrient water diatoms against the salinity ofthe water concentrations either remained high (in the in which they were collected during the course case of silicate) or at least did not approach of our sampling in April, July, and September unmeasurably low levels after the volcanic (Figure 13). All species declined from fresh to eruption (Figures 4, top and bottom and 5, brackish water, some more dramatically than top), and chlorophyll a concentrations re others . Most of the species disappeared from mained high in May 1980 (Figure 7); hence, water with salinities > 5%0 regardless of sea these variables were not responsible for the son, and in July three of the five species were large decrease in primary production immedi drastically reduced at 2.5%0. Diatoms do not ately after the eruption. Furthermore, two trib make up the complete phytoplankton assem utary rivers (region 9) draining into Youngs blage in either fresh or brackish water, yet Bay from the south were not affected by the they undoubtedly demonstrate the general large turbid flows entering the Columbia River trend for many freshwater cells to rupture and upstream of our study area, and so they re lose their chlorophyll on encountering even tained their high combined areal production very low salinities (Morris et al. 1978, 1982). rate in May 1980 (Figure 12). Youngs Bay The large decreases in both chlorophyll a and itself (region 4), however, did experience re production between regions 7 and (3 + 5) are duced primary production as a result of tur thus mainly the result of freshwater forms bidity-induced decrease in the photic depth. encountering the low-salinity tidal excursions Finall y, primary production rebounded by that reach mid-estuary in summer. In winter, July to rates more typical of summer rates, chlorophyll biomass and primary production Primary Production in the Columbia River Estuary I-LARA-LARA ET AL. 33

I Asterionella formosa 37.~~ I 2 Cyelotella glomerato 250t~ 3 cyctotetto stel ligera 4 Diatoma tenue 5 Fragilaria cr ot onens is C\I- 6 Melosira i slandie a o 7 Me/os ia i t aliea X 10.0 8 Na vicu!a cryptocephata 9 Sfephanodiseus astraea E 2 en 9 2 0::: 7.5 W CD ~ ~ Jul y 1980 Sep t.1980 Z 5.0

-l -l W u 2.5

2 4 o 42 6 o 2 4 SAL I NITY (%0)

FIGURE 13. Abundances offreshwater diatom species as a function ofsalinity in the Columbia River estuary, at three different times of year.

are generally low as a result of reduced light primary production in the Zaire River, de intensity and day length, and river flow is creasing values in the estuarine waters, and higher than in late summer. These effects tend increasing values again in the plume waters at to damp biomass and production differences sea. Similar patterns have been reported in the between contiguous regions along the estu Ems estuary in the Wadden Sea (Cadee and arine axis (Figures 7 and 9). Filardo and Dun Hegeman 1974), in the Rhone River estuary stan (1985) have observed these same effects (Blanc et al. 1969),in the upper part ofthe St. in the James River estuary, Morris et al. (1978, Lawrence estuary (Cardinal and Therriault 1982) in the Tamar estuary, and Cloern et al. 1976, Cote and Lacroix 1979), and in some (1983) in northern San Francisco Bay. branches ofthe Mississippi River (Riley 1937, The pattern ofhigh riverine production and Thomas and Simmons 1960). These patterns biomass, followed by reduced production and contrast with those of other estuaries, many biomass in the uppe r mixing zone at very low from the eastern seaboard of the United States, salinities, and terminating in slightly higher in which the estuary proper supports higher production and biomass at the estuary mou th phytoplankton biomass and production than in summer, has been observed in other estu either the major entering river(s) or the adja arine systems. Cadee (1978), for examp le, re cent coastal ocean. Along the Pacific coast, ported maximum values ofchlorophyll a and the Fraser River estuary seems to be an exam-

' ,"'U!" 34 PACIFIC SCIENCE, Volume 44, January 1990

TABLE 5

PHYTOPL ANKTON PRIMARY PRODUcnON IN SOME NORTH AMERICAN EsTUAR IES

ESTUARY REFERENCES

Columbia River estuary, Oregon 100 This study Fraser River estuary, British Columb ia 120 Parsons et al. (1970) Bedford Basin, Nova Scotia 220 Platt (1975) St. Margaret' s Bay, Nova Scotia 190 Platt and Conover (1971) Narragansett Bay, Rhod e Island 310 Furnas et al. (1976) Chesapeake Bay (upper) 125-510 Biggs and Flemer (1972) Chesapeake Bay (middle) 450-570 Stross and Stottlemeyer (1965) Chesapeake Bay (lower) 385 Fourni er (1966) Neuse River estuary , North Carol ina 300-500 Fisher et al. (1983) South River estuary, North Carolina 300-500 Fisher et al. (1983) Pamlico River estuary, North Carolina 200-500 Nixon et al. (1986) North Inlet, South Caro lina 260 Nixon et al. (1986) Burrad Inlet, British Columbia 350 Nixon et al. (1986) pIe in which the estuarine water is more pro lumbia estuary is also strongly suggested by ductive than its source waters (Parsons et al. the fact that primary production and chlo 1970, Takahashi et al. 1973). The estuarine rophyll a concentrations were only slightly portion ofSan Francisco Bay may be an inter higher in the shallows and bay s than in the mediate case, with phytoplankton biomass as main axis, but not statistically significant. In abundant in the riverine sources as in the some systems , for example in the northern estua ry, but decreasing sharply seaward ofthe reach ofSan Francisco Bay (Cloern 1979) and estuary. in the Po tomac River estuary (DiToro et al. The biomass and production pattern ex 1977), shallow regions are sites of ra pid phy hibited by the Columbia River estuary likely toplankton population increase. This simpl y results from the short residence time ofwater does not occur in the shallower areas of the in the estuary. A residence time of2 to 5 days Columbia estu ary , indicating the short resi (Neal 1965, Jay and Smith [in press]) suggests dence times ofcells in these areas as well as in that freshwater phytoplankton species are the main axis. rapidly carried into the low-salinity zone in A generally light-limited, short-residence mid-estuary, with little time for adjustment to time system in which the phytoplankton com mixing conditions. Even a tiny salinity change, munity is also subject to salinity-induced de when imposed abruptly on cells being trans pletions within the system should be a rela ported rapidly through the system , apparently tively nonproductive system . Indeed, when can effect immediate and dramatic changes to compared to a sampling ofother North Amer the phytoplankton community. It is perhaps ican estuaries, primary production in the Co important to note that the water residence lumbia estuary is the lowest (Table 5). The time in the Zaire estuary is also 2 or 3 days Columbia estuary is quite productive in terms (Eisma and van Bennekom 1978), while resi ofbenthic infauna and other invertebrate life, dence times in the Delaware estuary and in however (Jones et al. [in press]). Autotrophic Narragansett Bay, for example, are respec production within the estuary therefore is tively about 3 months (B. H. Ketchum, un likely not the primary, direct supplier of food published report on the distribution ofsalinity rations for sma ll invertebrate org anisms in the in the estuary of the Delaware River, Woods estuary. Rather it is more likely that secondary Hole Oceanographic Insti tution, Woods Hole , production is dependent upon large amounts Mass., 1952) and I month (Kremer and Nixon of particulate organic matter imported into 1978). the estua ry from upriver. Indeed, Bristow et Rapid transport of cells through the Co- al. (1985) determined from a laser fluoro sensor Primary Production in the Columbia River Estuary I -LARA-LARA ET AL. 35 survey of surface chlorophyll along a 734-km rophyll, dissolved organics and optical segment ofthe lower Snake and Columbia riv attenuation. Int. J. Remote Sensing 6: ers in late Ju ne 1982 that chlorophyll concen 1707-1734. trations were highest (>20 mg m-3) in a long CADEE, G . C. 1978. Primary production and stretch of river just upstream from our study chlorophyll in the Zaire River, estuary and area. Evaluation of particulate transport into plume. Neth. J. Sea Res. 3/4 :368-381. and through the estuary is the subject of a CADEE, G. C., and J. H EGEMAN. 1974. Primary second paper. production of phytoplankton in the D utch Wadden Sea. Neth. J. Sea Res. 8: 240-259. CARDINAL, A., and L. B. THERRIAULT. 1976. Le phytoplancton de l'estuarie moyen de Saint-Laurent en amont de L'Ile-Aux ACKNOWLEDGMENTS Coudres (Quebec). Int. Rev. Gesampten Sandy Moore and Rae Deane Leatham Hidrobiol. Hydrogr. 61 :639-648. provided inva luable assistance in parts of the CLOERN, J. E. 1979. Phytoplankton ecology of field program and in some laboratory analyses. the San Francisco Bay system :The status of our current understanding. Pages 247-264 in 1 . J. Conomos, ed. San Francisco Bay: The urbanized estuary. Pacific Division, LITERATURE CITED American Association for the Advancement of Science, San Francisco, California. AMSPOKER, M . C., and C. D . McINTIRE. 1986. CLOERN, J. E., A. E. ALPINE, B. E. COLE, R. L. Effects ofsedimentary processes and salinity J. WONG, J. F. ARTHUR, and M. D. BALL. on the diatom flora of the Columbia River 1983. River discharge controls phytoplank estuary. Bot. Mar. 24 :391-399. ton dynamics in the northern San Francisco ATLAS, E. L., S. W. HAGER, L. I. GORDON, and Bay estuary. Estuarine Coastal Shelf Sci. P. K. PARK. 1971. A practical manual for 16:415- 429. the use of the Technicon Autoanalyzer in COTE, R. , and G. LACROIX. 1979. Influence de seawater nutrient analysis: revised. Techni debits eleves et variables d'eau douce sur Ie cal Report 215, School of Oceanography, regime saisonnier de production primaire Oregon State University, Corvallis. d'un fjord subarctique. OceanoI. Acta 2 : BALL, M. D. , and J. F. ARTHUR. 1979. 299-306. Planktonic chlorophyll dynamics in the DIToRO, D. M. , R. V. THOMAN, D . J. northern San Francisco Bay and delta. O'CONNOR, and J. L. MANCINI. 1977. Estu Pages 265-285 in T. J. Conomos, ed. San arine phytoplankton biomass models. Veri Francisco Bay: The urbanized estuary. fication analyses and preliminary applica Pacific Division, American Association for tions. Pages 969-1020 in E. D . Goldberg, the Advancement ofScience, San Francisco, I. N. McCave, and J. H . Steele, eds. The sea. California. Vol. 6. John Wiley and Sons, New York. BIGGS, R. B., and D. A. FLEMER. 1972. The EISMA, D. , and A. J. VAN BENNEKOM. 1978. flux of particulate carbon in an estuary. The Zaire river and estuary and the Zaire Mar. BioI. 12 : 11- 17. outflow in the Atlantic Ocean. Neth. J. Sea BLANC, F ., M. LEVEAU, and K. H. SZEKIELDA . Res. 12:255-272. 1969. Effects eutrophiques au debouche FILARDO, M. J., and W. M. DUNSTAN. 1985. d'un grand fleuve (Grand Rhone). Mar. Hydrodynamic control of phytoplankton BioI. 3: 233-242. in low salinity waters of the James River BRISTOW, M . P. F., D. H .BUNDY, C. M . estuary, Virginia, U.S.A. Estuarine Coastal EDMONDS, P. E. PONTO, B. E. FREY, and Shelf Sci. 21:653-668. L. F. SMALL. 1985. Airborne laser fluorosen FISHER, T. R., P. R. CARLSON, and R. T. sor survey of the Columbia and Snake BARBER. 1983.Carbon and nitrogen prim ary rivers: Simultaneous measurements ofchlo- productivity in three North Carolina 36 PACIFIC SCIENCE, Volume 44, January 1990

estuaries. Estuarine Coastal Shelf Sci. 15: MORRIS, A. W., A. J. BALE, and R. J. M. 621-644. HOWLAND. 1982. Chemical variability in the FOURNIER, R. O. 1966. Some implications of Tamar Estuary, Southwest England. nutrient enrichment on different temporal Estuarine Coastal Shelf Sci. 14: 649-661. stages of a phytoplankton community. MORRIS, A. W. , R. F. C. MANTURA, A. J. Chesapeake Sci. 7: 11-19. BALE, and R. J. M. HOWLAND. 1978. Very FREY, B. E., J. R. LARA-LARA, and L. F. low salinity regions ofestuaries; important SMALL. 1983. Reduced rates of primary sites for chemical and biological reactions. production in the Columbia River Estuary Nature (London) 274: 678-680. following the eruption of Mt. Saint Helens NEAL, V. T. 1965. A calculation of flushing on 18 May 1980. Estuarine Coastal Shelf times and pollution distribution for the Sci. 17:213-218. Columbia River estuary. Ph .D. dissertation, FURNAS, M. J., G . L. HITCHCOCK, and T. J. Oregon State University, Corvallis. SMAYDA. 1976. Nutrient-phytoplankton NIXON, S. W., C. A. OVIATT, J. FRITHSEN, and relationships in Narragansett Bay during B. SULLIVAN. 1986. Nutrients and the pro the 1974 summer bloom. Pages 118-133 in ductivity of estuarine and coastal marine M . L. Wiley, ed. Estuarine processes: Uses, ecosystems. J. Limnol. Soc. S. Afr. 12: 43 stresses and adaptation to the estuary. Vol. 71. 1. Academic Press, New York. PARK, P. K., M. CATALFOMO, G. R. WEBSTER, HAERTEL, L. S., and C. L. OSTERBERG. 1967. and B. H. REID. 1969. Nutrients and carbon Ecology ofzooplankton, benthos, and fishes dioxide in the Columbia River. Limnol. in the Columbia River Estuary. Ecology Oceangr. 14: 559-567. 48: 459-472. PARK, P. K., C. L. OSTERBERG, and W. O. HAERTEL, L. S., C. OSTERBERG, H. CURL, and FORSTER. 1972. Chemical budget of the P. K. PARK. 1969. Nutrient and plankton Columbia River. Pages 123-134 in A. A. ecology of the Columbia River Estuary. Pruterand D . L. Alverson, eds. The Colum Ecology 50 :962-978. bia River estuary and adjacent ocean wa HALPERN, D. 1976. Structure of a coastal ters . University of Washington Press, upwelling event observed off Oregon Seattle. during July 1973. Deep-Sea Res. 23 :495 PARSONS, T. R., R. J. LEBRASSEUR, and W. E. 508. BARRACLOUGH. 1970. Levels of production JAY, D . A., and J. D . SMITH. In press. in the pelagic environment of the Strait of Circulation, density distribution and neap Georgia, British Columbia: A review. J. spring transitions in the Columbia River Fish. Res. Board Can. 27: 1251-1264. Estuary. Prog. Oceanogr. PENNOCK, J. R. 1985. Chlorophyll distribu JONES, K. K., O. A. SIMENSTAD, D . L. HIGLEY, tions in the Delaware estuary: Regulation and D. L. BOTTOM. In press. Community by light-limitation. Estuarine Coastal Shelf structure, distribution, and standing stock Sci. 21 :711-726. ofbenthos, epibenthos, and plankton in the PETERSON, R. E. 1977. A study of suspended Columbia River Estuary. Prog. Oceanogr. particulate matter: Arctic Ocean and north KREMER, J. N ., and S. W. NIXON. 1978. A ern Oregon continental shelf. Ph.D. disser coastal marine ecosystem. Simulation and tation, Oregon State University, Corvallis. analysis. Springer-Verlag, New York. PLATT, T. 1975. Analysis ofthe importance of LARA-LARA, J. R. 1982. Primary biomass and spatial and temporal heterogeneity in the production processes in the Columbia River estimation of the annual production by estuary. Ph.D. dissertation, Oregon State phytoplankton in a small, enriched, marine University, Corvallis. basin. J. Exp. Mar. BioI. Ecol. 18:99-109. MACISAAC, J. J., and R. C. DUGDALE. 1972. PLATT, T., and R. J. CONOVER. 1971. Variabil Interactions oflight and inorganic nitrogen ity and its effects on the chlorophyll budget in controlling nitrogen uptake in the sea. of a small marine basin. Mar. BioI. 10: 52 Deep-Sea Res. 19:209-232. 65. Primary Production in the Columbia River Estuary I-LARA-LARA ET AL. 37

RILEY, G. A. 1937. The significance of the THOMAS, W. H., and E. G. SIMMONS. 1960. Mississippi River drainage for biological Phytoplankton production in the Missis conditions in the northern Gulf of Mexico. sippi delta. Pages 103-116 in F. P. Shepard, J. Mar. Res. I: 60-74. ed. Recent sediments, Northwest Gulf of ROWE, K. , and R. BRENNE. 1981. Statistical Mexico. American Association of Petro interactive programming system (SIPS). leum Geologists, Tulsa, Oklahoma. Statistical Computing Report no. 7. Oregon VINCENT, W. F. 1981. Photosyntheticcapacity State University, Corvallis. measured by DCMU induced chlorophyll RYTHER, J. H. , and W. M. DUNSTAN. 1971. fluorescence in an oligotrophic lake. Fresh Nitrogen, phosphorus, and eutrophication water BioI. II :61-78. in the coastal marine environment. Science VOLLENWEIDER , R. A. 1971. A manual on 171: 1008 ~1O13. methods for measuring primary production SAMUELSON, G., and G. OQUIST. 1977. A in aquatic environments. IBP Handbook method for studying photosynthetic capac no . 12, Blackwell Scientific Publications, ities of unicellular algae based on in vivo Oxfo rd. chlorophyll fluorescence. Physioi. Planta WALSH, J. J., T. E. WHITLEDGE, J. C. KELLEY, rum 40: 315-319. S. A. HUNTSMAN, and R. D. PILLSBURY. SHARP, J. H., C. H. CULBERSON, and T. M . 1977. Further transition states of the Baja CHURCH. 1982. The chemistry of the California upwelling ecosystem. Limnoi. Delaware estuary. General considerations. Oceanogr. 22: 264-280. Limnoi. Oceanogr. 27: 1015-1028. WETZEL, R. G ., and G. E. LIKENS. 1979. SHERWOOD, C. R., and J. S. CREAGER. In press . Limnological analyses. W. B. Saunders, Sedimentary geology ofthe Columbia River Philadelphia, Pennsylvania. Estuary. Prog. Oceanogr. WILLIAMS, L. G. 1964. Possible relationships SIMENSTAD, C. A., L. F. SMALL, C. D. between plankton diatom species numbers McINTIRE, D. A. JAY, and C. R. SHERWOOD. and water-quality estimates. Ecology In press. An introduction to the Columbia 45: 809-823. River Estuary: Brief history, prior studies, 1972. Plankton diatom species and the role ofthe CREDDP studies. Prog. biomasses and the quality of American Oceanogr. rivers and the Great Lakes. Ecology 53 : SMALL, L. F., and D . W. MENZIES. 1981. Pat 1038-1050. terns of primary productivity and biomass WILLIAMS, L. G. , and C. SCOTT. 1962. Princi in a coastal upwelling region. Deep-Sea Res. pal diatoms of major waterways of the 28: 123-149. United States. Limnoi. Oceanogr. 7: 365 STRICKLAND, J. D. H., and T. R. PARSONS. 375. 1972. A practical handbook of seawater WILLIAMS, R. B. 1973. Nutrient levels and analysis. J. Fish. Res. Board Can. Bull. 167. phytoplankton productivity in the estuary. STROSS, R. G., and J. R. STOTTLEMEYER. 1965. Pages 59-89 in R. H. Chabreck, ed. Pro Primary production in the Patuxent River. ceedings ofthe Coastal Marsh and Estuary Chesapeake Sci. 6: 126-140. Management Symposium. Louisiana State TAKAHASHI, M. , K. FUJII, and T. R. PARSONS. Uni versity, Baton Rouge. 1973. Simulation study of phytoplankton photosynthesis and growth in the Fraser River estuary. Mar. BioI. 19:102-116.